Read on Twitter
I keep hearing doctors argue that the FDA should ensure drugs are 'safe', but should not be in the game of assessing if they are 'effective'
Let's leave that to the market.
I fear this stance reveals a lack of understanding of safety [thread]
Let's leave that to the market.
I fear this stance reveals a lack of understanding of safety [thread]
There are 2 ways we understand drug safety
The first and very limited way is to run a phase 1 trial, and ascertain at what dose level do drugs have significant adverse events.
A phase 1 trials can give you a sense of the Maximal Tolerated Dose, beyond which a drug is v unsafe
The first and very limited way is to run a phase 1 trial, and ascertain at what dose level do drugs have significant adverse events.
A phase 1 trials can give you a sense of the Maximal Tolerated Dose, beyond which a drug is v unsafe
But merely being below the MTD is not an assurance a drug is 'safe', and often it is only very large data sets, such as large phase 3, randomized trials or even larger real world data sets that fully characterize safety profiles
Real world datasets have a bad time of adjudicating safety if the adverse outcome also commonly occurs naturally (e.g. death) and a much easier time if the adverse signal is really unusual
e.g. https://jamanetwork.com/journals/jama/fullarticle/187878
e.g. https://jamanetwork.com/journals/jama/fullarticle/187878
What this essentially means is that there is no methodological way to ensure safety within the range of what reasonable people would think is safe for common endpoints in the absence of large rigorous studies that also assess efficacy
So it is really non-sensical to contend the role of a federal agency is to assess safety not efficacy, instead you have 2 choices
a. assess both
b. assess neither
a. assess both
b. assess neither
I favor assessing both, but I reject the idea that there can be a 'safety' assessment without 'efficacy' in any meaningful sense of the terms
